listening to music and physiological and psychological
TRANSCRIPT
Listening to music and physiological and psychological functioning:
The mediating role of emotion regulation and stress reactivity
M.V. Thomaa, U. Scholzb, U. Ehlertc and U.M. Naterd*
aDepartment of Psychology, Brandeis University, Brown 129, MS 062, 415 South Street,Waltham, MA 02453 2728, USA; bSocial and Health Psychology, University of Zurich,
Binzmuehlestrasse 14, Box 26, Zurich 8050, Switzerland; cClinical Psychology andPsychotherapy, University of Zurich, Binzmuehlestrasse 14, Box 26, Zurich 8050,Switzerland; dClinical Biopsychology, University of Marburg, Gutenbergstrasse 18,
Marburg 35032, Germany
Music listening has been suggested to have short term beneficial effects.The aim of this study was to investigate the association and potentialmediating mechanisms between various aspects of habitual music listeningbehaviour and physiological and psychological functioning. An internetbased survey was conducted in university students, measuring habitualmusic listening behaviour, emotion regulation, stress reactivity, as well asphysiological and psychological functioning. A total of 1230 individuals(mean 24.89� 5.34 years, 55.3% women) completed the questionnaire.Quantitative aspects of habitual music listening behaviour, i.e. averageduration of music listening and subjective relevance of music, were notassociated with physiological and psychological functioning. In contrast,qualitative aspects, i.e. reasons for listening (especially ‘reducing lonelinessand aggression’, and ‘arousing or intensifying specific emotions’)were significantly related to physiological and psychological functioning(all p 0.001). These direct effects were mediated by distress augmentingemotion regulation and individual stress reactivity. The habitual musiclistening behaviour appears to be a multifaceted behaviour that is furtherinfluenced by dispositions that are usually not related to music listening.Consequently, habitual music listening behaviour is not obviously linkedto physiological and psychological functioning.
Keywords: music; physiological and psychological functioning; distressaugmenting emotion regulation; individual stress reactivity
Introduction
The development of effective stress prevention or management has become an
important endeavour of current research efforts in health psychology. In particular,
cost-efficient interventions, such as music therapy, have received special interest in
terms of coping with stress and stress-related health issues. The potential of music to
trigger intensive emotions (Blood & Zatorre, 2001) and to attenuate psychophys-
iological activation (Khalfa, Bella, Roy, Peretz, & Lupien, 2003; Nater, Abbruzzese,
*Corresponding author. Email: nater@uni marburg.de
Krebs, & Ehlert, 2006) might be the main reason for its beneficial effect and,
consequently, for its increased therapeutic use in health care (see reviews in Evans,
2002; Nilsson, 2008).
Previous research has investigated the beneficial effect of listening to music in the
context of medical settings (Ebneshahidi & Mohseni, 2008; Pothoulaki et al., 2008;
Voss et al., 2004) and in laboratory-based studies (Khalfa et al., 2003; Roy, Peretz,
& Rainville, 2007). These studies found that listening to music has an acute,
i.e. short-term beneficial effect on various psychological and physiological
parameters in the listener. In addition, these short-term effects seem to vary
depending on the type of music (Burns et al., 2002; Roy et al., 2007).
Listening to music on a regular basis can be understood as habituation to the
possible effects of music. From this perspective, the question arises of whether short-
term beneficial effects accumulate over time, i.e. are there any long-term beneficial
effects of listening to music? Long-term effects have mainly been studied in the
context of formal music training and potential enhancement of non-musical aspects
of cognitive development (for a review see Schellenberg, 2001). Research into
beneficial long-term effects of listening to music on physiological and psychological
functioning has largely been neglected, with a few notable exceptions. In animals,
for instance, repeated music exposure significantly altered neurotrophin levels in the
hypothalamus (Angelucci, Ricci, Padua, Sabino, & Tonali, 2007), which may alter
hypothalamus–pituitary–adrenal axis activation. In humans, repeated music listen-
ing resulted in the following significant benefits: less depressed and confused mood
in patients after stroke (Sarkamo et al., 2008), reductions in high systolic blood
pressure (Teng, Wong, & Zhang, 2007), better quality of sleep in elderly patients
(Lai & Good, 2005) and reduced depression, mood disturbances, fatigue, and
decreases in cortisol concentration in healthy adults (McKinney, Antoni, Kumar,
Tims, & McCabe, 1997). Repeated everyday listening to music may therefore also
have a beneficial long-term effect.
Barely any attempts have been undertaken to reveal the underlying mechanisms
or variables that mediate the effect of repeated music listening on various aspects
of physiological and psychological functioning. Mitchell, MacDonald, Knussen, and
Serpell (2007) conducted a cross-sectional survey with chronic pain sufferers in which
the authors assessed music-listening behaviour and quality of life (QOL) as well as
pain levels. Their findings indicate that specific aspects of music-listening behaviour,
such as the average time per day spent listening to music, subjective relevance
of music and reasons for listening to music (RML), may play an essential role.
The latter were all concerned with pain-related coping (e.g. ‘to help pain’, ‘to relieve
tension and stress’, ‘to feel relaxed’, ‘to express emotions’). However, it is to be
expected that a variety of other reasons for listening are relevant in a non-patient
sample. For example, in the sample by North and colleagues, healthy adolescents’
reasons for listening differed quite widely, e.g. ‘to relieve boredom’, ‘to help me
through difficult times’, ‘to please friends’, ‘to reduce loneliness’ (North, Hargreaves,
& O’Neill, 2000). However, these authors did not examine associations between
reasons for listening and indicators for physiological and psychological functioning
in their participants. Thus, to the best of our knowledge, there are no data regarding
the relationship between RML and indicators for physiological and psychological
functioning in healthy adults.
Moreover, Mitchell et al. (2007) provided no explanation regarding the
underlying mechanisms that may mediate the purported association between reasons
228
for music listening and physiological and psychological functioning. Previous
research indicates that music listening is extensively used for emotion regulation
purposes (Juslin & Sloboda, 2010; Saarikallio & Erikkila, 2007; Thoma, Ryf, Ehlert,
& Nater, 2006). Emotion regulation is defined as ‘. . . a process by which individuals
influence which emotions they have, when they have them, and how they experience
and express these emotions’ (Gross, 1998, p. 275). It has been found that individuals
differ in their habitual way to regulate their emotions (i.e. dispositional emotion
regulation; Mohiyeddini, 2005; see also Wirtz et al., 2006). What is more, it has been
reported that emotion regulation is linked to psychological and physiological
functioning (Lam, Dickerson, Zoccola, & Zaldivar, 2009). However, potential
mediating effects of emotion regulation processes on the association between
listening to music and physiological and psychological functioning have not been
described in the literature so far. RML have also been related to stress-reducing
purposes (Thayer, Newman, & McCain, 1994). The term ‘stress reduction’ refers to
techniques or strategies that aim to decrease the experience of stress which occurs
whenever situational demands tax or exceed the resources of an afflicted individual
(Lazarus & Folkman, 1984). As per emotion regulation, individuals differ with
regard to their general reactivity to stress (Schulz, Jansen, & Schlotz, 2005), which
has been related to psychological and physiological functioning. Therefore, it might
be assumed that, depending on how individuals habitually respond to stressors
(i.e. individual stress reactivity), people’s RML, and consequently, its effect on
physiological and psychological functioning, may differ. However, as with emotion
regulation, no mediator model has tested an assumed mediating effect of individual
stress reactivity on the association between listening to music and physiological and
psychological functioning.
In the current research literature on the beneficial effect of listening to music, we
identified a lack of studies investigating associations between various aspects of
habitual music-listening behaviour and physiological and psychological functioning
in adults. Potential mediating mechanisms, such as emotion regulation processes
or individual stress reactivity, have not yet been examined. We therefore set out to
investigate the interrelation between various aspects of habitual music-listening
behaviour and physiological and psychological functioning, and the potential
mediating influence of emotion regulation processes and individual stress reactivity.
Method
Participants and data collection procedure
A large online survey in a student population at the University of Zurich and the
Zurich Federal Institute of Technology was conducted. Both institutions provided
electronic mailing lists of students and faculty members who had previously provided
permission of being contacted for research purposes. The email recipients were sent
an invitation email with the invitation to enrol online for a survey about music, body
and emotion. In the case of interest, the potential participants had access to the
survey via an anonymous internet link embedded in the email. In order to ensure
higher participation rates (Bosnjak & Tuten, 2003), six cinema tickets and a dining
coupon were raffled off. The survey was programmed online using the software
Unipark (Unipark, Germany). For statistical analysis, the collected data were
exported into SPSS 17.0 (SPSS, Chicago, Illinois). In a pilot study, a minimum time
229
requirement of 25min to complete the questionnaire was determined; participants
who answered all questions substantially more quickly than others were excluded
from the data set. All participants took part in the survey voluntarily and provided
informed consent. The study was approved by the Swiss ethics committee of the
Canton of Zurich.
Materials
The complete online survey consisted of 292 items. The first section consisted of the
informed consent and questions addressing sociodemographic and general health-
related variables. Participants who did not provide informed consent were unable to
continue with the questionnaire and automatically directed to exit the questionnaire.
The second section comprised questions about habitual music-listening behaviour
(average duration of music listening per day, reasons and motivations for listening to
music (RML), preferred music styles and subjective relevance of music), emotion
regulation (dispositional emotion regulation), stress variables (dispositional stress
reactivity) and health variables (somatic complaints and QOL). The following
questionnaires were presented in random order.
Music-listening behaviour and music preference
The Music Preference Questionnaire (Nater, Krebs, & Ehlert, 2005) was used to
assess music-listening behaviour in terms of subjective importance of music, time
spent listening to music and RML, as well as general music preference. As RMLs,
the following 10 options were given: to relax, be activated, be distracted, reduce
aggression, reduce loneliness, arouse specific emotions, increase specific emotions,
work better and avoid boredom. The respondents could additionally add their own
reasons under ‘others’.
Physiological and psychological functioning
In the current study, physiological and psychological functioning was operationalised
by the German self-report ‘Freiburg Complaint List’ (Freiburger Beschwerdeliste,
FBL; Fahrenberg, 1994) and a German version of the Quality of Life Questionnaire
of the World Health Organization (WHOQoL-5, WHOQoL-Group, 1998a, b). The
FBL is a well-validated inventory for different aspects of bodily complaints and
normal functioning. The questionnaire consists of nine subscales (General
Condition, Tiredness, Cardiovascular, Gastrointestinal, Head-Throat, Tenseness,
Emotional Reactivity, Pain and Sensory) and a summary scale, which are derived
from 80 items. Fifty-seven items, such as ‘Do you get headaches?’, have to be
answered by indicating one of five frequency categories (almost everyday, about
three times a week, about twice a month, about twice a year and almost never).
Twenty-three items, such as ‘Are you sensitive to pain?’, have to be answered by
indicating one of five intensity categories (very strongly, strongly, moderately, barely
and insensitive). Higher scores on FBL scales indicate frequent and intense
complaints and go along with a negative general well-being. The FBL has been
standardised with a representative sample of N¼ 2070 German women and men
aged 16 and older (Fahrenberg, 1995). Reliability coefficients (Cronbach’s alpha) for
the subscales and the total score range between 0.73 and 0.90. In terms of validity,
230
the total score of the FBL has been shown to be substantially associated with
frequency of visiting practitioners and of in-patient treatment over and above effects
of gender, age and social class. Moreover, FBL scores were consistently higher
for psychotherapy patients compared to controls (Fahrenberg, 1995).
The WHOQoL-instruments (WHOQoL-100 and WHOQoL-BREF, a 25 items
version of the WHOQoL; Angermeyer, Kilian, &Matschinger, 2000) generally assess
participants’ QOL ‘. . .a broad-ranging concept, incorporating in a complex way
the person’s physical health, psychological state, level of independence, social
relationships, and their relationship to salient features of their environment’
(WHOQoL-Group, 1994, p. 43). The WHOQoL-instruments have been standardised
cross-culturally using samples from several countries among them a German-
speaking version. The WHOQoL-instruments have been demonstrated to have good
discriminate validity with regard to ill and well respondents; good content validity
as well as internal consistency (Cronbach’s alpha ranging from 0.66 to 0.84 for the
subscales; e.g. WHOQoL; Group, 1998). In the current study, we applied the
WHOQoL-5. The WHOQoL-5 is a short version of the WHOQoL-instruments and
consists of five items covering overall life satisfaction. Participants are asked to
answer the items with regard to the last 2 weeks. Six answer options range from 0
(at no time point) to 5 (the whole time). The WHOQoL-5 has not yet been tested for
its psychometric measures in healthy participants, but has previously been applied
in various different WHO projects (Biering-Sorensen et al., 2006).
Emotion regulation
The German Inventory for Regulation of Emotion (ERI, [Inventar zur
Emotionsregulationswirksamkeit]; Mohiyeddini, 2005; see also Wirtz et al., 2006)
was used to assess the dispositional regulation of emotions. The ERI has 34 items
and comprises a hedonistic (HED) way of emotion regulation, a distress-augmenting
regulation (DAR), i.e. intensifying of negative emotions, and an emotional
moderation regulation (MOD), i.e. buffering of emotions. Using a 6-point Likert
scale ranging from 1 (hardly ever) to 6 (almost always), the participants were asked
to rate how often they use specific strategies (e.g. go out with friends, thinking of
something cheerful) to regulate their emotions.
Inter-individual stress reactivity
The German Stress Reactivity Scale (SRS, [Stress-Reaktivitaets-Skala]; Schulz et al.,
2005) was used to assess general susceptibility to stress, i.e. the disposition to respond
to stressors with intense, immediate and long-lasting stress reactions. Four
personality characteristics are assumed to determine stress reactivity: high intrusive-
ness, low self-efficacy, high arousability of the central nervous system and high
negative affectivity. The SRS assesses general and specific stress reactivity in
different stressor domains.
Data analyses
Data were analysed by means of regression analyses using path models with AMOS
17.0 (Arbuckle & Wothke, 1999). Mediator hypotheses were tested according to
recommendations by Baron and Kenny (1986). For each full model, three regression
231
analyses were conducted. The first tested whether the predictor was significantly
associated with the outcome. The second tested whether the predictor was
significantly associated with possible mediators and the final analysis included
the predictor and the mediator variables as predictors of the outcome. When the
previously significant direct effect of the predictor on the outcome decreases to
a non-significant level after the mediator variables are included, the effect can be
regarded as fully mediated. Reduction in strength of the association can be seen as
a partial mediation. Furthermore, the indirect effect from the predictor through the
mediator on the outcome was tested for statistical significance (i.e. whether or not it
is different from zero) using the Sobel test (Sobel, 1982).
Missing data were treated using the full information maximum likelihood
(FIML) algorithm. FIML is a method that uses all available information of all
observed (also incomplete) cases in order to enhance the validity of parameter
estimation. FIML is superior to common methods such as listwise or pairwise
deletion (Enders & Bandalos, 2001) and was thus chosen for the present studies.
Univariate and multivariate outliers were routinely screened and treated as suggested
by Tabachnick and Fidell (2001).
Results
Participants’ characteristics
A total of N¼ 22,179 recipients were sent an invitation email of which N¼ 2616
provided data sets. This is a response rate of 11.79%. After cleaning the data pool,
i.e. removing incomplete or incorrect data sets, and excluding data sets of
participants who answered the questionnaire substantially faster than within 25
min, we obtained a final N of 1230 (24.89� 5.34 years old; 683 females (55.3%) and
547 males) participants who completed all questionnaires. All participants were
students or junior university staff members and were therefore highly comparable
with regard to their level of education.
Four models were tested: the first mediator model specified RML reducing
loneliness, the second RML reducing aggression as predictor, DAR and stress
reactivity as mediators and QOL and sum scores of the FBL as outcomes. The third
and fourth mediator models specified RML arousing specific feelings and RML
intensifying specific feelings as predictors, DAR and stress reactivity as mediators
and only the sum scores of the FBL as outcomes (a table of raw correlations between
variables of interest is available upon request).
Mediator model with RML reducing loneliness
In a first model, we tested whether there were direct effects of RML reducing
loneliness on QOL and the sum scores of the FBL when controlling for gender.
Results are displayed in Figure 1. The substantial associations between RML
reducing loneliness and QOL (�¼ÿ0.21) and sum scores of the FBL (�¼ 0.24)
remained when controlling for gender (�¼ 0.27 for the association with the sum
scores of the FBL, indicating higher reporting of symptoms of the FBL for women;
�¼ÿ0.01, ns for the association with QOL). The definite mediator model included
stress reactivity and DAR as mediator variables. Indirect effects between RML
reducing loneliness via stress reactivity to QOL (Sobel Z¼ÿ7.22, p¼ 0.001) and the
232
233
R2=0.06
R2=0.14
Figure I. Mediator model with RML reducing loneliness~ Notes: Coefficients in parentheses: direct effects without mediator variables. All coefficients reported are controlled for gender. Associations with gender: {3 0.09, with DAR; {3 0.29 with stress reactivity; {3 0.16 with quality of life; {3 0.14 Ylith sum scores of the FBL; r 0.05 with RML reducing loneliness. Gender coded 1 men, 2 women. Unless otherwise indicated, all coefficients were significant alp< 0.01; ap < 0.10.
sum scores of the FBL (Sobel Z= 7.39, p =0.001) as well as via DAR to sum scores of the FBL (Sobel Z=5.30, p=0.001) reached significance. The indirect effect between RML reducing loneliness via DAR to QO L (Sobel Z = - 1.82, p = 0.07) was not significant at the 5% level. Overall, due to the considerable reduction of the direct effects, the associations between RML reducing loneliness and QOL and sum scores of the FBL are mectiated by stress reactivity and for values of sum scores of the FBL as outcomes also in part by DAR.
Mediator mode/with RML reduci11g aggressio11
In a second model, we tested whether there were direct effects of RML reducing aggression on QOL and sum scores of the FBL when controlling for gender. The associations between RML reducing aggression and QOL (/3 = -0.15) and sum scores of the FBL (/3 =0.17) remained almost unchanged when controlling for gender (/3= 0.27 for the association with the sum scores of the FBL, indicating higher reporting of symptoms of the FBL for women; f3 = -0.01, ns for the association with QOL). The definite mediator model included stress reactivity and DAR as mectiator variables. ResuJts are displayed in Figure 2 and are comparable to RML reducing loneliness. Indirect effects between RML reducing aggression via stress reactivity to QOL (Sobel Z=-4.37, p=O.OOl) and sum scores of the FBL (Sobel Z=4.41, p=0.001) as well as via DAR to sum scores of the FBL (Sobel Z=4.73, p=O.OOl) reached statistical significance. The indirect effect between RML reducing aggression via DAR to QOL (Sobel Z = -1.95, p = 0.05) was not significant at the 5% level. Overall, due to the considerable reduction of the direct effects, the associations between RML reducing aggression and QOL and sum scores of the FBL were mectiated by stress reactivity and for sum scores of the FBL as outcomes also by DAR.
234
R2=0.11
Figure 2. Mediator model with RML reducing aggression. Notes: Coefficients in parentheses: direct effects without mediator variables. All coefficients reported are controlled for gender. Associations with gender: f3 0.1 0, with DAR; f3 0.29 with stress reactivity; f3 0.11 with quality of life; f3 0.14 with sum scores of the FBL; r 0.05, p < 0.10 with RML reducing aggressioiL Gender coded I men, 2 women. Unless otherwise indicated, all coefficients were significant at p < 0.01; np < 0.05.
Mediato1· mode/with RM L arousing spec~fic feelings
In a third model, we tested whether there were direct effects of RML arousing specific feelings on the sum scores of the FBL when controlling for gender. The associations between RML arousing specific feelings and sum scores of the FBL (.8 = 0.16) remained unchanged when controlling for gender (.8 = 0.28 for the association with the sum scores of the FBL, indicating higher reporting of symptoms in the FBL for women; .B = - 0.01). The definite mediator model included stress reactivity and DAR as mediator variables. Results are displayed in Figure 3. Both indirect effects between RML arousing specific feelings via stress reactivity to sum scores of the FBL (Sobel Z = 3.54, p =0.001) as well as via DAR to sum scores of the FBL (Sobel 2=5.69, p=O.OOl) reached significance. Overall, due to the reduction of the direct effect, the association between RML arousing specific feelings and sum scores of the FBL was mediated by stress reactivity and by DAR.
Mediator mode/with RJ\1/L intensifying spec(fic feelings
In a fourth model, we tested whether there were direct effects of RML intensifying specific feelings on sum scores of the FBL when controlling for gender. The associations between RML intensifying specific feelings and sum scores of the FBL (.8 = 0.19) remained unchanged when controlling for gender (8 = 0.29 for the association with sum scores of the FBL, indicating higher reporting of symptoms in the FBL for women). The definite mediator model included stress reactivity and DAR as mediator variables. As results are very similar to those for RML arousing specific feelings, no additional figure is provided. RML intensifying specific feelings was associated with .8=0.10 (p=O.OOl) to stress reactivity and with .8=0.32 (p= 0.001) with DAR. Stress reactivity and DAR, in turn, were significantly related
0.28
R·=0.09
Distressaugmenting regulation
0.18
l---- 0.07a [0.16)-----4>1
I 0.21
0.39
R·=0.10
Figure 3. Mediator model with RML arousing specific feelings.
235
R2=0.30
Notes: Coefficient in parentheses: direct effect without mediator variables. All coefficients reported are controlled for gender. Associations with gender: {3 0.11, with DAR; {3 0.30 with stress reactivity; {3 0.14 with sum scores of the FBL; r 0.01 , ns with RML arousing specific feelings. Gender coded I men, 2 women. Unless otherwise indicated, all coeffi cients were significant at p < 0.01; ap < 0.05.
to sum scores of the FBL (~=0.39, p= O.OOI for stress reactivity and ~= 0.16,
p =0.001 for DAR). Both indirect effects between RML intensifying specific feelings via stress reactivity to sum scores of the FBL (Sobel Z = 3.39, p = 0.001) as well as via DAR to sum scores of the FBL (Sobel 2 = 5.51, p=O.OOl) reached significance. Overall, due to the reduction of the direct effect, the association between RML intensifying specific feelings and sum scores of the FBL was mediated by stress rea.ctivi ly and by DAR.
Discussion
This study examined the interrelation between habitual music-listening behaviour and physiological and psychological functioning. We hypothesised that emotion regulation processes and individual stress reactivity accounted for potential associations between habitual music-listening behaviour and indicators of physiological and psychological functioning. Our findings indicate that average duration of music listening or subjective relevance of music are not associated with physiological and psychological functioning. In contrast, RML (especially 'reducing loneliness and aggression' and 'arousing or intensifying specific emotions') seem to play a much more crucial role. These effects were mediated by distress-augmenting emotion regulation and individual stress reactivity.
It appears that in non-patient adults, qualitative aspects of listening to music, i.e. reasons related to emotion regulation, are more important for the general physiological and psychological functioning than quantitative aspects, i.e. average duration of music listening or subjective relevance of music. This result is in contrast to fmdings by Mitchell et al. (2007), who found that both frequent music listening and subjective importance of music were related to higher QOL and lesser need for medical treatment in pain sufferers. In our study, the mean ratings of subjective
relevance of music were rather high in most participants. Consequently, the fact that
we did not find an interrelation between the subjective relevance of music and
indicators of physiological and psychological functioning may be due to a ceiling
effect. Moreover, given the particular characteristics of the sample by Mitchell et al.
(2007), i.e. chronic pain sufferers, comparisons with our sample, i.e. non-patient
adults, are limited. Differences in the basic health status of the two samples
are probably the main reason for the identified discrepancies between the results
of the two studies.
We found a direct negative association between the reason for listening ‘to
reduce loneliness’ and indicators of physiological and psychological functioning
(i.e. negative association with QOL and positive association with number of somatic
symptoms). Other studies examining chronic pain sufferers (Mitchell et al., 2007) and
adolescents (North et al., 2000) also reported that reducing loneliness was a major
reason for listening to music. However, the negative association with indicators
of physiological and psychological functioning has not yet been documented in the
literature. Although we did not measure loneliness per se, we might assume that
those respondents who reported listening to music in order to reduce loneliness may
also experience this feeling more often. One experiences the distressing feeling of
loneliness in times when one’s own social contacts are perceived as less fulfilling than
one whishes them to be (Peplau & Perlman, 1982). Loneliness and its related social
isolation have been associated with increased physical health risks (Cole, 2008),
reduced longevity (Peel, McClure, & Bartlett, 2005), increased risk of cancer
(Reynolds & Kaplan, 1990), increased depressive symptoms (Cacioppo & Hawkley,
2003) and poor lifestyle behaviour, such as smoking, physical inactivity and
overweight (Hawkley, Thisted, & Cacioppo, 2009; Lauder, Mummery, Jones, &
Caperchione, 2006). Our finding of a negative association between the reason of
listening to reduce loneliness, as an indirect indicator of loneliness, and indicators
of physiological and psychological functioning is therefore in line with previous
studies relating to the topic of ill health through loneliness.
Further, we found a direct negative association between the reason for listening
‘to reduce aggression’ and indicators of physiological and psychological functioning
(i.e. negative association with QOL and positive association with number of somatic
symptoms). This is a novel finding, which has not been reported previously. As with
loneliness, we did not measure aggression as such, but we may assume that those
respondents reporting to listening to music in order to reduce (feelings or
behavioural tendencies associated with) aggression may also experience it more
often. Aggression describes a behaviour that is performed to injure or hurt somebody
who avoids being a target of these actions (Baron & Richardson, 2004). It has
previously been associated with several health issues, such as depression and chronic
stress, or poor lifestyle behaviour, such as smoking and drinking (Johnson, 1990).
Our finding of a negative association between the reason of listening to reduce
aggression (as an indirect indicator of aggression) and indicators of physiological
and psychological functioning is therefore consistent with previous findings relating
to the topic of ill health through aggression.
Finally, we found a direct negative association between the reasons for listening
‘arousing and intensifying specific feelings’ and indicators of physiological and
psychological functioning (i.e. positive association with number of somatic symp-
toms). This is in line with previous studies (Mitchell et al., 2007; North &
Hargreaves, 2000). The emotion-eliciting capacity of music is considered to be the
236
main reason for actively engaging with music (Juslin & Sloboda, 2001). Accordingly,
our finding appears rather surprising and raises new questions. This observation can
only be interpreted when the mediating variables, i.e. DAR and individual stress
reactivity, are taken into account. DAR describes the disposition of an individual
to intensify negative emotions. It is therefore assumed that individuals who have
the disposition to intensify negative emotions also listen to music with the purpose of
arousing or intensifying negative emotions. Previous research indicates that negative
emotions are associated with negative health outcomes (Kubzansky & Kawachi,
2000). Our finding is therefore consistent with results of previous investigations.
The same explanation might also be valid for the second mediating variable,
i.e. stress reactivity. The construct of stress reactivity describes the disposition of an
individual to respond to stressors with immediate, strong and long-term stress
reactions (Schulz et al., 2005). We therefore assume that respondents with this
disposition also experience emotions associated with stress more often (such as
anxiety or anger) and consequently preferentially listen to music with the purpose
of arousing or intensifying these emotions. Sustained stress experience and its
biological consequences negatively impact psychological and physiological health
(Esch, Stefano, Fricchione, & Benson, 2002). Thus, the negative association between
the reason for listening ‘arousing and intensifying specific feelings’ and indicators of
physiological and psychological functioning can be explained by the mediating
influence of stress reactivity.
Does listening to music improve physiological and psychological functioning?
Our findings indicate that it might not be possible to answer this question with a
simple yes or no. Habitual music-listening behaviour appears to be a multifaceted
behaviour that consists of many different aspects. In addition, the use of music seems
to be further influenced by individual dispositions (emotion regulation and
susceptibility to stress). Moreover, it is likely that use and subsequent effects of
music listening are further influenced by other variables, such as personality factors
(Chamorro-Premuzic, Swami, Furnham, & Maakip, 2009). Finally, physiological
and psychological functioning is also further influenced by various habitual
behaviours, such as physical exercise (Waddington, Malcolm, & Green, 1997) or
diet (Cordain et al., 2005). From this perspective, the fact that we were still able to
detect some associations between habitual music-listening behaviour and indicators
of physiological and psychological functioning further emphasises the relevance
of this topic. However, to achieve a better understanding of this association, future
studies should also include the above-mentioned influencing factors.
There are several limitations to this study that should be acknowledged.
Survey respondents: As we were studying the associations between music listening
and indicators of physiological and psychological functioning in a wide university
population, the generalisation of our results beyond this sample is restricted. It is to
be assumed that participants in our sample, consisting mainly of university students,
have different everyday lives than other samples, and might therefore show different
music-listening behaviour. Moreover, as respondents were aware of the topic of the
survey, i.e. music, and due to the voluntary nature of study participation, a selection
bias may have confounded our results. The high amount of average time spent
listening to music and the high values of the subjective relevance of music are
indicators of this assumption. Consequently, the findings should be replicated in a
237
larger and more heterogeneous sample consisting of respondents who are more naıve
to the intentions of the survey.
Setting: A web-based data collection has many advantages: besides its convenience
and anonymity for respondents, it is cost-effective and allows access to a broad and
homogeneous sample as well as the assessment of data in a short period of time. The
shortcomings of this approach are the assumed biased, i.e. non-random response and
the generally lower response rates in comparison to conventional paper-and-pencil
methodology (for a review see Cook, Heath, & Thompson, 2000).
Cross-sectional design: Given the fact that our survey was a cross-sectional study, we
are only able to reveal associations between the variables of interest and unable to
establish any causal interpretations between them. Only a broadly constructed
prospective study design, i.e. a longitudinal correlational study, which controls for a
multitude of influencing variables, might provide a comprehensive answer to the
question of whether listening to music in everyday life does indeed have a positive
long-term impact on health.
In conclusion, this was the first study to investigate the associations between
music-listening behaviour and indicators of physiological and psychological func-
tioning in a sample of adult non-patients. We were able to demonstrate that
qualitative aspects of music-listening behaviour, i.e. specific (negative) emotion-
regulated reasons for listening, are directly associated with indicators of physiolog-
ical and psychological functioning. We also found that these associations are
mediated by dispositional emotion regulation processes and individual stress
reactivity. Thus, individuals use music differentially, depending on their habitual
way of regulating their emotions and on their individual susceptibility to stress.
Based on this knowledge, music listening might be more systematically used in
therapeutic settings, thus enhancing its clinical relevance.
Acknowledgements
This study was supported by a grant from the Young Investigator Grant of the Universityof Zurich. Grant no: 56233208 (MVT). We thank the survey respondents and acknowledgethe assistance of Martial Boni and Regula Flutsch. We would like to thank Prof. ChangizMohiyeddini for providing us with the ERI. UMN acknowledges support by the VolkswagenFoundation.
References
Angelucci, F., Ricci, E., Padua, L., Sabino, A., & Tonali, P.A. (2007). Music exposure
differentially alters the levels of brain derived neurotrophic factor and nerve growth factor
in the mouse hypothalamus. Neuroscience Letters, 429, 152 155, doi:10.1016/
j.neulet.2007.1010.1005.
Angermeyer, M.C., Kilian, R., & Matschinger, H. (2000). WHOQOL 100 und WHOQOL
BREF: Handbuch fur die deutsche Version der WHO Instrumente zur Erfassung von
Lebensqualitat. Gottingen: Hogrefe.
Arbuckle, J.L., & Wothke, W. (1999). Amos (version 4.0). Chicago: Smallwaters.
238
Baron, R.M., & Kenny, D.A. (1986). The mediator moderator variable distinction in social
psychological research: Conceptual, strategic, and statistical considerations. Journal of
Personality and Social Psychology, 51, 1173 1182.
Baron, R.A., & Richardson, D.R. (2004). Human aggression (2nd ed.). New York, NY:
Plenum.
Biering Sorensen, F., Scheuringer, M., Baumberger, M., Charlifue, S.W., Post, M.W.,
Montero, F., & Stucki, G. (2006). Developing core sets for persons with spinal cord
injuries based on the international classification of functioning, disability and health as a
way to specify functioning. Spinal Cord, 44, 541 546.
Blood, A.J., & Zatorre, R.J. (2001). Intensely pleasurable responses to music correlate with
activity in brain regions implicated in reward and emotion. Proceedings of the National
Academy of Sciences of the United States of America, 98, 11818 11823.
Bosnjak, M., & Tuten, T.L. (2003). Prepaid and promised incentives in web surveys:
An experiment. Social Science Computer Review, 21, 208 217.
Burns, J.L., Labbe, E., Arke, B., Capeless, K., Cooksey, B., Steadman, A., & Gonzales, C.
(2002). The effects of different types of music on perceived and physiological measures
of stress. Journal of Music Therapy, 39, 101 116.
Cacioppo, J.T., & Hawkley, L.C. (2003). Social isolation and health, with an emphasis
on underlying mechanisms. Perspectives in Biology and Medicine, 46, 39 52.
Chamorro Premuzic, T., Swami, V., Furnham, A., & Maakip, I. (2009). The big five
personality traits and uses of music. Journal of Individual Differences, 30, 20 27.
Cole, W.C. (2008). Social regulation of leukocyte homeostasis: The role of glucocorticoid
sensitivity. Brain, Behavior, and Immunity, 22, 1049 1055.
Cook, C., Heath, F., & Thompson, R.L. (2000). A meta analysis of response rates in web or
internet based surveys. Education and Psychological Measurement, 60, 821 836.
Cordain, L., Eaton, S.B., Sevastian, A., Mann, N., Lindeberg, S., Watkins, B.A., &
Brand Miller, J. (2005). Origins and evolution of the Western diet: Health implications
for the 21th century. American Journal of Clinical Nutrition, 81, 341 354.
Ebneshahidi, A., & Mohseni, M. (2008). The effect of patient selected music on early
postoperative pain, anxiety, and hemodynamic profile in cesarean section surgery.
Journal of Alternative Complementary Medicine, 14, 827 831.
Enders, C.K., & Bandalos, D.L. (2001). The relative performance of full information
maximum likelihood estimation for missing data in structural equation models.
Structural Equation Modeling, 8, 430 457.
Esch, T., Stefano, G.B., Fricchione, G.L., & Benson, H. (2002). An overview of stress and
its impact in immunological diseases. Modern Aspects of Immunobiology, 2, 187 192.
Evans, D. (2002). The effectiveness of music as an intervention for hospital patients:
A systematic review. Journal of Advanced Nursing, 37, 8 18.
Fahrenberg, J. (1994). Die freiburger beschwerdenliste (FBL): Form FBL G und revidierte form
FBL R. Gottingen: Hogrefe.
Fahrenberg, J. (1995). Somatic complaints in the German population. Journal of
Psychosomatic Research, 39, 809 817.
Gross, J.J. (1998). The emerging field of emotion regulation: An integrative review. Review of
General Psychology, 2, 271 299.
Group, T.W. (1998). The World Health Organization Quality of Life Assessment (WHOQoL):
Development and general psychometric properties. Social Science and Medicine, 46,
1569 1585.
Hawkley, L.C., Thisted, R.A., & Cacioppo, J.T. (2009). Loneliness predicts reduced physical
activity: Cross sectional & longitudinal analysis. Health Psychology, 28, 354 363.
Johnson, E.H. (1990). The deadly emotions: The role of anger, hostility, and aggression in health
and emotional well being. New York, NY, England: Praeger.
Juslin, P.N., & Sloboda, J.A. (2001). Music and emotion: Theory and research. London:
Oxford University Press.
239
Juslin, P.N., & Sloboda, J. (2010). Handbook of music and emotion. New York, NY: Oxford
University Press.
Khalfa, S., Bella, S.D., Roy, M., Peretz, I., & Lupien, S.J. (2003). Effects of relaxing music
on salivary cortisol level after psychological stress. Annals of the New York Academy of
Sciences, 999, 374 376.
Kubzansky, L.D., & Kawachi, I. (2000). Going to the heart of the matter: Do negative
emotions cause coronary heart disease? Journal of Psychosomatic Research, 48, 323 337.
Lai, H.L., & Good, M. (2005). Music improves sleep quality in older adults. Journal of
Advanced Nursing, 49, 234 244.
Lam, S., Dickerson, S.S., Zoccola, P.M., & Zaldivar, F. (2009). Emotion regulation and
cortisol reactivity to a social evaluative speech task. Psychoneuroendocrinology, 34,
1355 1362.
Lauder, W., Mummery, K., Jones, M., & Caperchione, C. (2006). A comparison of health
behaviours in lonely and non lonely populations. Psychology, Health and Medicine, 11,
233 245.
Lazarus, R.S., & Folkman, S. (1984). Stress, appraisal, and coping. New York, NY: Springer.
McKinney, C.H., Antoni, M.H., Kumar, M., Tims, F.C., & McCabe, P.M. (1997). Effects
of guided imagery and music (GIM) therapy on mood and cortisol in healthy adults.
Health Psychology, 16, 390 400.
Mitchell, L.A., MacDonald, R.A.R., Knussen, C., & Serpell, M.G. (2007). A survey
investigation of the effects of music listening on chronic pain. Psychology of Music, 35,
37 57.
Mohiyeddini, C. (2005). Personality, emotion regulation and psychopathology.University of
Tuebingen, Germany.
Nater, U.M., Abbruzzese, E., Krebs, M., & Ehlert, U. (2006). Sex differences in emotional
and psychophysiological responses to musical stimuli. International Journal of
Psychophysiology, 62, 300 308.
Nater, U.M., Krebs, M., & Ehlert, U. (2005). Sensation seeking, music preference and
psychophysiological reactivity to music. Musicae Scientiae, 9, 239 254.
Nilsson, U. (2008). The anxiety and pain reducing effects of music interventions: A systematic
review. AORN Journal, 87, 780 807.
North, A.C., & Hargreaves, D.J. (2000). Musical preferences during and after relaxation and
exercise. American Journal of Psychology, 113, 43 67.
North, A.C., Hargreaves, D.J., & O’Neill, S.A. (2000). The importance of music to
adolescents. The British Journal of Educational Psychology, 70, 255 272.
Peel, N.M., McClure, R.J., & Bartlett, H.P. (2005). Behavioral determinants of healthy aging.
American Journal of Preventive Medicine, 28, 298 304.
Peplau, L.A., & Perlman, D. (1982). Perspectives on loneliness. In L.A. Peplau &
D. Perlman (Eds.), Loneliness: A sourcebook of current theory, research and therapy
(pp. 1 20). New York, NY: Wiley.
Pothoulaki, M., MacDonald, R.A.R., Flowers, P., Stamataki, E., Filiopoulos, V.,
Stamatiadis, D., & Stathakis, Ch.P. (2008). An investigation of the effects of music on
anxiety and pain perception in patients undergoing haemodialysis treatment. Journal of
Health Psychology, 13, 912 920.
Reynolds, P., & Kaplan, G.A. (1990). Social connections and risk for cancer: Prospective
evidence from the Alameda County Study. Behavioral Medicine, 16, 101 110.
Roy, M., Peretz, I., & Rainville, P. (2007). Emotional valence contributes to music induced
analgesia. Pain, 134, 140 147.
Saarikallio, S., & Erikkila, J. (2007). The role of music in adolescents’ mood regulation.
Psychology of Music, 35, 88 109.
Sarkamo, T., Tervaniemi, M., Laitinen, S., Forsblom, A., Soinila, S., Mikkonen, M., Autti,
T., & Hietanen, M. (2008). Music listening enhances cognitive recovery and mood after
middle cerebral artery stroke. Brain, 131, 866 876.
240
Schellenberg, E.G. (2001). Music and nonmusical abilities. Annals of the New York Academy
of Sciences, 930, 355 371.
Schulz, P., Jansen, L., & Schlotz, W. (2005). Stressreaktivitat: Theoretisches Konzept und
Messung. [Stress reactivity: Theoretical concept and measurement]. Diagnostica, 51,
124 133.
Sobel, M.E. (1982). Asymptotic intervals for indirect effects in structural equations models.
In S. Leinhart (Ed.), Socio meth (Vol. 1982, pp. 290 312). San Francisco, CA: Jossey Bass.
Tabachnick, B.G., & Fidell, L.S. (2001). Using multivariate statistics (4th ed.). Boston, MA:
Allyn and Bacon.
Teng, X.F., Wong, M.Y.M., & Zhang, Y.T. (2007). The effect on music on hypertensive
patients. In Proceedings of the 29th Annual International Conference of the IEEE
EMBS (pp. 4649 4651). Lyon: Chinese University of Hong Kong. doi:10.1109/
IEMBS.2007.4353376.
Thayer, R.E., Newman, J.R., & McCain, T.C. (1994). Self regulation of mood: Strategies for
changing a bad mood, raising energy and reducing tension. Journal of Personality and
Social Psychology, 67, 910 925.
Thoma, M.V., Ryf, S., Ehlert, U., & Nater, U.M. (2006). Regulation of emotions by listening
to music in emotional situations. In M. Baroni, A.R. Addessi, R. Caterina, & M. Costa
(Eds.), Proceedings of the 9th International Conference on Music Perception and Cognition.
Italy: University of Bologana.
Voss, J.A., Good, M., Yates, B., Baun, M.M., Thompson, A., & Hertzog, M. (2004). Sedative
music reduces anxiety and pain during chair rest after open heart surgery. Pain, 112,
197 203.
Waddington, I., Malcolm, D., & Green, K. (1997). Sport, health and physical education:
A reconsideration. European Physical Education Review, 3, 165 182.
WHOQoL Group (1994). The development of the World Health Organization Quality of
Life Assessment Instrument (WHOQOL). In J. Orley & W. Kuyken (Eds.), Quality of life
assessment: International perspectives. Heidelberg: Springer Verlag.
WHOQoL Group (1998a). Development of the World Health Organization WHOQOL
BREF quality of life assessment. The WHOQOL Group. Psychological Medicine, 28,
551 558.
WHOQoL Group (1998b). The World Health Organization Quality of Life Assessment
(WHOQOL): Development and general psychometric properties. Social Science and
Medicine, 46, 1569 1585.
Wirtz, P., Von Kanel, R., Mohiyeddini, C., Emini, L., Ruedisueli, K., Groessbauer, S., &
Ehlert, U. (2006). Low social support and poor emotion regulation are associated
with increased stress hormone reactivity to mental stress in systemic hypertension.
Journal of Clinical Endocrinology and Metabolism, 91, 3857 3865.
241